The proposed research will investigate the nature of the implicit models people use to construct, understand and evaluate social relationships. The research focuses on the ways in which four models structure social interaction in a wide variety of domains: work, decision making, women's relationships, marriage systems, socialization and the formation of new social relationships, moral standards and the definition of transgression, as well as interpretation of the moral significance of misfortune. The project will study the formal mathematical structures which map each of the four types of social relationship, and the distinctive mode of symbolic representation in which each model characteristically is encoded. The proposal shows how our studies of these domains complement each other to yield a synergistic understanding of the fundamental models which underlie them all. The basic method will be ethnographic participant observation and interviewing among the Moose in Burkina Faso, West Africa. Microcomputer text search and retrieval software will aid in the derivation of reliable and unbiased inductive inferences from the qualitative ethnographic data. To complement this ethnographic approach, the project includes interviews with victims of a variety of misfortunes (including AIDS) in both Burkina Faso and the U.S. The plan also includes a learning, memory, and communication experiment to investigate the modes of symbolic encoding of the four models, as well as a field interview procedure for testing what kinds of quantitative transformations leave the qualitative structure of each type of social relationship intact. The study promises to clarify how people meet each other's social expectations and moral standards to constitute mutually intelligible and fulfilling relationships. So it should shed indirect light to help elucidate a wide range of factors related to mental health, particularly in regard to the conditions for effective and constructively coordinated psychosocial functioning.